Question

我使用Keras构建了一个自定义模型，并将其转换为TFLITE模型。 CNN模型的输入形状为60 * 60 * 3，因此数据类型为int时，所需的缓冲区为60 * 60 * 3 * 4 = 43200。现在，使用功能-

将模型与 Flutter 集成在一起

    recognitions = await Tflite.runModelOnImage(
          path: widget.imagePath,
          imageMean: 127.5,
          imageStd: 127.5,
          numResults: 2,
          threshold: 0.1,
          asynch: true);
      print(recognitions);
    }

我收到此错误-

    E/AndroidRuntime(21412): Caused by: java.lang.IllegalArgumentException: Cannot convert between a TensorFlowLite buffer with 21600 bytes and a Java Buffer with 43200 bytes.
E/AndroidRuntime(21412):    at org.tensorflow.lite.Tensor.throwIfShapeIsIncompatible(Tensor.java:332)
E/AndroidRuntime(21412):    at org.tensorflow.lite.Tensor.throwIfDataIsIncompatible(Tensor.java:305)
E/AndroidRuntime(21412):    at org.tensorflow.lite.Tensor.setTo(Tensor.java:123)
E/AndroidRuntime(21412):    at org.tensorflow.lite.NativeInterpreterWrapper.run(NativeInterpreterWrapper.java:150)
E/AndroidRuntime(21412):    at org.tensorflow.lite.Interpreter.runForMultipleInputsOutputs(Interpreter.java:311)
E/AndroidRuntime(21412):    at org.tensorflow.lite.Interpreter.run(Interpreter.java:272)
E/AndroidRuntime(21412):    at sq.flutter.tflite.TflitePlugin$RunModelOnImage.runTflite(TflitePlugin.java:452)
E/AndroidRuntime(21412):    at sq.flutter.tflite.TflitePlugin$TfliteTask.doInBackground(TflitePlugin.java:419)
E/AndroidRuntime(21412):    at sq.flutter.tflite.TflitePlugin$TfliteTask.doInBackground(TflitePlugin.java:393)
E/AndroidRuntime(21412):    at android.os.AsyncTask$2.call(AsyncTask.java:333)
E/AndroidRuntime(21412):    at java.util.concurrent.FutureTask.run(FutureTask.java:266)
E/AndroidRuntime(21412):    ... 4 more

关于如何解决此错误的任何建议？

使用tflite插件将TFLITE自定义模型与Flutter集成时遇到问题

0 个答案: